Odd Lots

AI Can Tell Us Something About Credit Market Weakness

December 4, 2025

Key Takeaways Copied to clipboard!

  • The increasing prevalence of specific structural protections in credit and M&A deal terms, tracked by Noetica, signals growing anxiety and a "flight to fortification" among lenders and borrowers in the market. 
  • AI-powered analysis of deal documents allows for the quantification of market standards for complex terms, such as erroneous payment clauses and J. Crew blockers, which are rapidly evolving in response to market events. 
  • Highly leveraged, circular financing structures seen in large AI infrastructure deals, like the Meta/Blue Owl transaction, involve underwriting immature assets with debt levels (90% leverage) that are exceptionally high by traditional LBO standards. 

Segments

AI Elevating American Workers
Copied to clipboard!
(00:00:00)
  • Key Takeaway: AI should elevate, not eliminate, workers by freeing them from drudgery.
  • Summary: An advertisement segment promoting Palantir’s AI vision, focusing on how it helps workers unlock potential in factories and hospitals.
Credit Market Weakness and AI
Copied to clipboard!
(00:01:48)
  • Key Takeaway: There are many credit-related issues, including private credit concerns and the potential application of AI to complex deal text.
  • Summary: Hosts Joe Weisenthal and Tracy Alloway welcome the guest and frame the discussion around current credit market concerns and the utility of LLMs in analyzing lengthy legal agreements.
Introducing Noetica AI and Deal Terms
Copied to clipboard!
(00:04:52)
  • Key Takeaway: Noetica AI uses AI to benchmark real-time data on market standards for credit and M&A deal terms.
  • Summary: Dan Wertman introduces his company, Noetica AI, which helps transactional professionals determine if their agreement terms are on or off market by benchmarking against comps.
The Importance of Deal Terms
Copied to clipboard!
(00:08:07)
  • Key Takeaway: Deal terms are the ‘rules of the road’ in finance, and missing or incorrect terms can have massive financial consequences, as shown by the Revlon case.
  • Summary: Wertman explains deal terms using a lease analogy and the famous McNugget/Revlon payment error example, highlighting the market adoption of ’erroneous payment deal terms’.
Quantifying Structural Fortification Trends
Copied to clipboard!
(00:11:15)
  • Key Takeaway: Lenders are increasingly demanding structural protections, signaling anxiety about potential distress.
  • Summary: Wertman details the ‘flight to fortification,’ citing massive quarterly increases in terms like anti-pet smart terms and anti-serta protections, which govern collateral security and payment order.
Borrower Flexibility and Leverage Ratios
Copied to clipboard!
(00:13:23)
  • Key Takeaway: Borrowers are gaining economic flexibility (e.g., EBITDA add-backs) in exchange for lender fortifications, while required leverage ratios are tightening.
  • Summary: The discussion covers borrower concessions like cost-savings add-backs and the tightening of required leverage ratios from 3.9x to 3.5x EBITDA.
AI Technology Stack and Data
Copied to clipboard!
(00:27:35)
  • Key Takeaway: Noetica uses proprietary, fine-tuned, open-source language models deployed in secure environments, not public APIs like ChatGPT, to build a knowledge graph of over a billion deal terms.
  • Summary: Wertman describes the technology stack, emphasizing security for sensitive documents and the use of semantic meaning extraction to track terms regardless of phrasing.
Lenders Preparing for Distress
Copied to clipboard!
(00:30:37)
  • Key Takeaway: The obsession with lien subordination terms (who gets paid first) suggests creditors are actively preparing for bankruptcy recovery, not just preventing liability management.
  • Summary: Wertman analyzes the sharp rise in lien subordination terms, comparing the current focus on recovery to adding insurance after locks and alarms are installed.
Off-Balance Sheet Financing in AI Deals
Copied to clipboard!
(00:38:35)
  • Key Takeaway: Complex, highly leveraged (90%) circular financing structures are being used to fund assets like AI data centers while keeping debt off the main corporate balance sheet.
  • Summary: Using a pizza analogy, Wertman explains the Meta/Blue Owl deal structure, highlighting the risk of underwriting immature assets with high leverage and contingent guarantees.
AI Competition and Market Winners
Copied to clipboard!
(00:44:20)
  • Key Takeaway: AI is a generational technology shift, but the race to define the category will result in many headline losers, even if the ultimate winners are massive.
  • Summary: Wertman discusses the existential nature of AI competition and how fast-moving markets necessitate rigorous attention to deal documentation, citing the Frank/JPM fraud case as an example of overlooked details.